EBola Series The Capiz Times Point & Counterpoint September 3, 2014 10:14 pm Ebola Hits Capiz part 1 of three-part series Virgilio Clavel While Pres. Noynoy Aquino was hurling political tirades to his detractors, including Cong. Manny “Pacman” Pacquiao, confusion gripped Capisnons when news that Ebola had hit Capiz several days ago. Ebola stands for Electoral Bola (Joke). Dr. Edwin Chinel Monares recently came here to clarify the confusion on the survey results conducted by the Global Zenith Research and Consulting’s, Inc. (GZRCI) for the period July 19-20, 2014. He is the chairman and CEO of GZRCI. GZRCI survey was done to determine the (awareness, 2) trust, 3) satisfaction, 4) and preference, or ATS & P, of Capiz’s provincial political personalities. Its scope covers 16 municipalities and one city, with 1203 respondents, using a stratified random and proportional sampling, at 97% confidence level and 3 percent margin of error (MoE). The stratum sample sizes for the 1st and 2nd electoral districts are 613 (50.96%) and 590 (49.04 %). respondents respectively. The Sloven Formula was the sampling design technique used (note: as to its propriety or statistical validity, discussion follows below). Inclusion of “survey item personalities” was done by subscription, in consultation with the politician who commissioned the said survey. Tables 12 and 13 list the following 19 political personalities namely, in alphabetical order, Andaya (Mary), Berjamin (Beboy), Bermejo (Dante), Bermejo (Vic), Billones (Tawi), Castro (Oto), Celino (Allan), Contreras (Dodoy), Contreras (Nonoy), Cruz-Am (Fritz), Del Rosario (Tony), Escutin (Bobo), Labao (Leodegario), Longares (Joel), Martinez (Bulilit), Orosco (Turic), Palomar )Karen), Patricio (Ike), and Tanco (Vic). The “EBola conspiracy” could not have stirred popular notice were it not for Felipe “Ipe” Celino’s information flash in local print and social media. Ipe, writing also for Philippine Daily Inquirer and Panay News, “skewed a little upward through back door source” the results on Table 15 of the July 2014 GZRCI survey. DYJJ Capiz shared this same information flash post in the internet. In short, the sources of the non-GZRCI version came from Ipe Celino and DYJJ Capiz. In managing counter-propaganda, the better approach is to always cite Ipe Celino and DYJJ Capiz, say according to x x x – the author of the rumor or media spin. Monares, in an interview, disowned Ipe’s Table 15 version which put del Rosario on top spot instead of Vice Gov. Nonoy Contreras. He claimed Ipe Celino’s version did not contain his signature, thus fake. The correct ranking results, he stressed, are: 1) Nonoy Contreras (53.95%, with 649 fN), followed by 2) Dodoy Contreras (14.63%, with fN 176), 3) Joel Longares (10.97%, with 132 fN) (see also Nonoy Contreras tops survey for Capiz governor, Panay News, August 16, 2014 issue; or Jean Benedicto’s post in the Capisnon Facebook Community). Whether it was Ipe Celino or someone else who twisted the GZRCI result – and committing as it was done – has no professional or ethical bearing on my part and for those expert statistician. Let me explain the EBola propaganda spew (my apology for using this term) in terms of 1) technical-fitness, and 2) common sense-speak. Jeffrey Tejada and Joyce Raymond Punzalan, on the Misuse of the Slovin’s Formula, wrote --- “Slovin’s Formula is applicable only when estimating a population proportion and when the confidence coefficient is 95%... It is optimal only when the population is suspected to be close to 0.05. Hence, it is not advisable to use Slovin’s Formula if any of the … assumptions do not hold … the formula (can) be used only when the assumptions are met…” ( In Philippine Statistician, Vol. 61, No. 1 (2012), pp. 129-136; see also Glenn Israel. 2013. Determining Sample Size. edis.ifas.ufl.edul; Cochran, W. 1977. Sampling Techniques, 3rd. ed. New York. John Wiley and Sons, Inc.; Yamane, T. 1967. Elementary Sampling Theory . New Jersey. Prentice-Hall, Inc.; Kish, L. 1965. Survey Sampling. New York. John Wiley and Sons, Inc.; ellinahs.blogspot/2006/01/surveys_slovin_formula.html.; Griford, J.P. and Fruchor B. 1973. Fundamental Statistics in Pstcholigy and Education. New York. MCGraw-Hill; Krebbs-Smith, Kott PS and Guenther PM. Mean Proportions and Population Proportion: Two Answers to the Same Question. J’Am Diet Assoc. 1989 May: 89(5): 671-6; Reforma, Mila. 1998. PM 299.1: Quantitative Methods in Public Policy and Administration. Quezon City. UP Open University). Given its survey’s confidence level at 0.97 and MoE at 0.03, GZRCI misapplied the Slovin (or Sloven) Formula considering that the basic assumptions were not present, i.e., at 0.95% confidence coefficient and at 0.5 MoE. (See David, I and Maligalig D. 2006. Are We Teaching Staistics Correctly to Our Youth? The Philippine Statistician, 55 (3 & 4): 1-28; and Pagoso, C and Moantana, R. 1995. Introductory Statistics. Manila. Rex Book Store). The GZRCI survey therefore is not valid and not reliable. It could have estimated a different sampling frame, using , more likely a different statistical tool, not necessarily a Slovin Formula. A sample proportion is different from a population proportion. A Slovin Formula is used to calculate normally the sample size proportion. Granting that Ipe Celino or his “bossing” mangled “behind upward” (or “e-skewed back”) the GZRCI result, no professional damage had been done because the result was ab initio invalid and unreliable. The mangling may figuratively be termed “an error based on error,” and thus lacks technical ascendancy or respect. The reliability, i.e., the consistency of independent but comparable measures of the same individual, group or situation (Reforma 1998), of the GZRCI result is technically diminished by the following: 1) Lack of explanatory assumption on how 457 unemployed respondents (Table 11) by magical stroke came out with Php 17,500 monthly average income-and-below (Table 7). The number 457 constitutes 43.14% of 976 respondents under this income category, which is 81.13% of the total 1203 survey population size. (to be continued)
Posted on: Sat, 13 Sep 2014 11:17:23 +0000